Annealing effects on the magnetic and structural properties of single ...

Physics

Physics fields Okayama University

Year 2000

Annealing effects on the magnetic and structural properties of single-crystal TDAE-C-60 Takashi Kambe∗

Yoshio Nogami†

Kokichi Oshima‡

∗ Department

of Physics, Faculty of Science, Okayama University of Physics, Faculty of Science, Okayama University ‡ Department of Physics, Faculty of Science, Okayama University This paper is posted at eScholarship@OUDIR : Okayama University Digital Information Repository. † Department

http://escholarship.lib.okayama-u.ac.jp/physics general/38

RAPID COMMUNICATIONS

PHYSICAL REVIEW B

VOLUME 61, NUMBER 2

1 JANUARY 2000-II

Annealing effects on the magnetic and structural properties of single-crystal TDAE-C60 Takashi Kambe, Yoshio Nogami, and Kokichi Oshima Department of Physics, Faculty of Science, Okayama University, 3-1-1 Tushimanaka, Okayama 700-8530, Japan 共Received 7 September 1999兲 Annealing effects on the magnetic and structural properties of single-crystal TDAE-C60 are investigated. When a crystal is well-annealed at 350 K, ferromagnetic ordering takes place below 16 K, though no magnetic phase transition is shown in as-grown crystal. The saturated magnetization was obtained to be 0.9⫾0.1␮ B per C60 . It was first found that the well-annealed crystal shows a structural phase transition around 180 K, probably associated with the orientational ordering of C60 molecules. On the other hand, the as-grown crystal undergoes no structural phase transition at least down to 30 K while the motion of C60 molecules is restricted below around 150 K. The possible relation between the low-temperature structure and the magnetic ordering is discussed.

The origin of the magnetic phase transition in TDAE-C60 has attracted significant attention. Although extensive studies have been carried out using powder samples,1 the ground state of this salt has not been settled. Recently, Blinc et al. succeeded in synthesizing single crystals2 and observed ferromagnetic resonance with an extremely small anisotropy field.3 They also reported4,5 that the ferromagnetic behavior was not observed unless the sample was well annealed above room temperature. While the ground state below T c is settled to be ferromagnet, both the annealing effects on the magnetic property and the microscopic origin of the ferromagnetic interactions in TDAE-C60 have remained as the subject of discussion. This ferromagnetic transition of 16 K is the highest among organic materials; therefore, the understanding of its mechanism is highly desirable. In this paper, we report effects of the annealing treatment on the magnetic and structural properties of single-crystal TDAE-C60 . Single crystals were prepared by directly mixing C60 solution in toluene with TDAE under a high-purity N2 gas condition. The solution was held in an incubator, whose temperature was kept about 27 °C, for two weeks. The typical size of the obtained crystal was about 1⫻0.5⫻0.5 mm3 . Since the present salts are unstable in air, the single crystals were put into a quartz capillary tube with He gas in order to prevent sample degradation. At first, we selected the as-grown single crystals without twin boundary using x-ray imaging plate diffractometer Rigaku R-AXISIV. Then, we measured the magnetization of the as-grown crystal using the MPMS2 superconducting quantum interference device 共SQUID兲 magnetometer. The crystal was annealed in this magnetometer at 350 K for 30 min and its magnetization was successively measured. We repeated this annealing procedure until its ferromagnetic magnetization below 16 K became independent of both temperature and magnetic field. After this annealing procedure, we checked the quality of the crystal at room temperature using x-ray. For low-temperature x-ray measurements of the as-grown crystals, the Huber four circle diffractometer 5042 with a monochromatized Mo K ¯␣ x-ray beam 共54 kV, 200 mA兲 and the Air Product DE201 cryocooler were used. For low-temperature x-ray measurements of the well-annealed 0163-1829/2000/61共2兲/862共4兲/$15.00

PRB 61

crystals, a homemade rotatable x-ray camera with a focused Mo K ¯␣ x-ray beam 共55 kV, 35 mA兲 and the Iwatani gas 501 cryocooler were used. The single crystal was mounted on the sample holder so as to rotate around the c axis. Electron spin resonance 共ESR兲 measurements were performed at X band using the Bruker ESP300e spectrometer equipped with the Oxford ESR910 continuous flow cryostat. We measured the temperature variations of ESR parameters 共spin susceptibility, linewidth, and g factor兲 of as-grown and well-annealed single crystals. Figure 1 shows the magnetization process of a single crystal at 5 K. The sample was cooled under zero-field conditions. For as-grown crystals, the susceptibility is Curie-like and no magnetic phase transition was detected below 5 K. When a crystal is first annealed, a ferromagnetic moment appears below around 16 K. In addition, the ferromagnetic component is enhanced with the passage of the annealing time and finally saturates. The time that is required for satu-

FIG. 1. Magnetization process of single-crystal TDAE-C60 at 5 K. The open and closed circles denote the results for the as-grown and the well-annealed crystals, respectively. The inset shows the temperature dependence of magnetization for the well-annealed crystal. R862

©2000 The American Physical Society


PRB 61

ANNEALING EFFECTS ON THE MAGNETIC AND . . .

R863

FIG. 3. Temperature variations of (⫺6,4,2) and (3,5,7) Bragg reflection intensities for the as-grown single-crystal TDAE-C60 .

FIG. 2. Temperature variations of ESR parameters 关共a兲 ␹ spin , 共b兲 g factor, and 共c兲 ⌬H pp 兴 of single-crystal TDAE-C60 . The open and closed circles denote the results for the as-grown and the wellannealed crystals, respectively. The magnetic field is applied parallel to the b axis.

rating the magnetization is about 180 min for the samples used. 共However, this value is not so meaningful because it probably depends on the sample mass.兲 Assuming that all unpaired spins are on each C60 molecule, the unpaired spin per C60 is estimated to be 0.9⫾0.1␮ B from the magnitude of saturated magnetization. Note that this saturated magnetization in the single crystal is significantly higher than those of 0.3␮ B in the powder annealed samples in the previous report.6 The temperature dependence of the magnetization for the well-annealed crystal is shown in the inset of Fig. 1. The ferromagnetic phase transition occurs around T c ⫽16 K, which is consistent with the previous result.2 The temperature variations of the ESR parameters above T c are shown in Fig. 2. For both as-grown and well-annealed crystals, the ESR line shape at room temperature corresponds neither to a Lorentzian nor a Gaussian, but the shape transforms to a Lorentzian below 150– 170 K. The spin susceptibility, ␹ spin , at room temperature remains at almost the same value even if the crystal is annealed. The ␹ spin for both cases exhibit Curie-like behavior in the whole temperature range with the same Curie constant. Thus, the amount of spin density, i.e., the amount of charge transfer from C60 to TDAE, is not affected by annealing. Our preliminary Raman-scattering measurements support this result. We observed A g (2) mode among C60 intermolecular vibrational modes located at 1463 cm⫺1 for both cases. Using the linear relation between

its frequency and the ionicity of C60 molecules,7 C60 molecules are considered to be monovalent. This value corresponds well to the results by magnetization and ESR measurements. Remarkable annealing effects were observed both for g factors and for linewidths (⌬H pp ) 关see Fig. 2共b兲 and 共c兲兴. In the case of as-grown crystals, ⌬H pp and g factors remain constant above 150 K for all directions. For T⬍150 K, ⌬H pp exhibits a sharp jump around 150 K with a hysteresis behavior and then the g factors begin to decrease below this temperature. In contrast, for a well-annealed one, ⌬H pp decreases gradually from room temperature to 20 K with cooling and has no jump around 150 K. The g factor in this case is almost independent of temperature through 150 K. Below 50 K, the steep increases of g factors for both cases are probably due to the demagnetization originated in the shape of crystals. Let us focus on the structural property of as-grown crystals. We investigated the space-group symmetry of as-grown crystals through the extinction rule of Bragg reflection. No substantial peak more than 10⫺4 times of the Bragg intensity were observed at the indices as follows: hkl: h⫹k⫽2n ⫹1, hk0: h⫹k⫽2n⫹1, h0l: h⫽2n⫹1 or l⫽2n⫹1, 0kl: k⫽2n⫹1, h00: h⫽2n⫹1, 0k0: k⫽2n⫹1, 00l: l⫽2n ⫹1, both at 30 K and 297 K. These results clearly show that the space group remains C2/c and no structural phase transition occurs. The tight-binding calculation indicates that ⫺1 deforms from the I h state to the D 3d , D 5d , and D 2h C60 states due to the Jahn-Teller effect.8 The thermal transition between these states is usually called a pseudorotation, because each state corresponds to one of different orientations ⫺1 molecule. NMR measurements in well-annealed of the C60 crystals indicate that this pseudorotation begins to freeze out below 170 K2. For as-grown crystals, the intensities of (⫺6,4,2) and (3,5,7) Bragg reflections show abrupt decreases above 150 K with heating 共see Fig. 3兲. These Bragg reflections have large diffraction angles; therefore, their intensities are dependent on temperature factors, i.e., the magTABLE I. Change of lattice parameters at room temperature.

as-grown well-annealed

a

b

c

␤

15.90共2兲 15.82共5兲

13.07共2兲 13.00共2兲

20.14共4兲 19.93共5兲

93.10 93.86


R864

TAKASHI KAMBE, YOSHIO NOGAMI, AND KOKICHI OSHIMA

PRB 61

FIG. 4. X-ray photographs for the well-annealed single-crystal TDAE-C60 . The photos 共a兲 and 共b兲 are taken at 200 and 40 K, respectively.

nitude of the thermal motion of atoms and molecules. The abrupt increases in the Bragg intensities in Fig. 3 suggest the restriction of C60 motion below 150 K. Since C60 is not a perfect sphere, it is natural that the orientational ordering of C60 leads to a symmetry change as seen in the pure C60 crystal.9 No symmetry change in the as-grown crystals indicates that C60 freezes without the perfect orientational order. The ESR line shape transforms to Lorentzian with the narrowing of linewidth below the freezing temperature. It results from the fact that the magnetic exchange interaction works efficiently between the orientationally frozen C60 molecules, as Blinc et al. proposed.2 The restriction of the C60 motion in the as-grown crystals is also reflected in the sharp decrease of the ESR linewidth below 150 K. Next, we check the change of lattice parameters at room temperature with annealing.6 Table I shows an example of the change of lattice parameters. It is found that the lengths of all the crystal axes shrink after annealing. Especially remarkable is the shrinkage of the c axis, along which the C60 molecules are closely packed 共about 1%兲, compared with the length before annealing. No change in the space group with annealing has been obtained. Next, we investigate the lowtemperature symmetry of the well-annealed crystals. X-ray patterns for the well-annealed crystal at 共a兲 200 K and 共b兲 40 K are displayed in Fig. 4, where the reflections in the reciprocal 关 a * ⫹b * 兴 c * plane are shown. The intense Bragg points along the longitudinal series have even h⫹k indices in Fig. 4共a兲. At 40 K, note the presence of new Bragg 共NB兲 reflections between these series. Since these NB reflections have odd h⫹k indices, the appearance of the NB reflection shows the structural phase transition from a C-centered lattice to a primitive lattice. An average NB reflection intensity is 0.1 times the average Bragg intensity. Figure 5 shows the typical temperature variations of the NB reflections. The smooth evolution of the NB reflection intensities with cooling indicates this phase transition is of second order. The temperature of the phase transition is around 180 K. The motion of the C60 molecules are continuously frozen below the same temperature range, because the ESR linewidth of the well-annealed crystals gradually decreases. Thus, the structural transition observed is probably associated with the orientational order of the C60 molecules. The relatively strong intensity of the NB reflections supports this idea. Since the phase transition is of second order, the space-group symmetry below 180 K will be a subgroup of C2/c. There are four subgroups of P2/c, P2/n, P2 1 /c, and P2 1 /n with the primitive lattice. The P2 1 /n space-group symmetry allows that the orientations of the C60 molecules are alternately

FIG. 5. Temperature variations of the normalized intensity of several new Bragg reflections for the well-annealed single-crystal TDAE-C60 .

ordered along the nearest (c axis兲 and next-nearest (a⫹b direction兲 neighbors. This orbital ordering structure is one of the models of the ferromagnetic interactions between C60 molecules.10 Finally, we discuss the relation between the magnetic ordering and the orientational ordering of C60 molecules. This coupling was also suggested by the high magnetic-field measurement.11 In the as-grown crystal, there is no structural phase transition, and C60 freezes in glassy state without the orientational order. The large hysteresis observed in the ESR linewidth seems to correspond to this orientational glass transition. The orientational disorder leads to the distribution of exchange interactions between molecules, resulting in no magnetic phase transition or in spin-glass behavior. On the other hand, in the well-annealed crystal, the second-order structural phase transition exists. The orientational order gradually grows below this temperature, and it is reflected in the continuous decrease of the ESR linewidth. As a result, the exchange interactions will become uniform and then the long-range ferromagnetic ordering occurs at T c . In summary, we found that the well-annealed TDAE-C60 , which has T c ⫽16 K, shows the structural phase transition around 180 K, probably associated with the orientational ordering of the C60 molecules. On the other hand, the as-grown crystal undergoes no structural phase transition at least down to 30 K, and possibly has the orientational disorder of the C60 molecules below 150 K. It is necessary to consider the ferromagnetic interactions in accordance with the lowtemperature structure. We plan to examine the lowtemperature x-ray measurements in well-annealed crystals. The authors would like to thank Professor K. Yakushi 共IMS, Okazaki National Research Organization兲 for the use of the Raman scattering system, the committee of the x-ray laboratory of Okayama University and that of the x-ray room in the Venture Business Laboratory of Okayama University for the use of their diffractometers, Dr. Y. Kubozono for the preparation of the C60 refinement, and K. Katayama and H. Negawa for experimental support. This research was supported by a Grant-in-Aid for Scientific Research from the Japanese Ministry of Education, Science, and Culture, and by the Okayama Foundation for Science and Technology.


PRB 61 1

ANNEALING EFFECTS ON THE MAGNETIC AND . . .

K. Tanaka, T. Sato, K. Yoshizawa, K. Okahara, T. Yamabe, and M. Tokumoto, Chem. Phys. Lett. 237, 271 共1995兲. 2 R. Blinc, K. Pokhodnia, P. Cevc, D. Arc˘on, A. Omerzu, D. Mihailovic´, P. Venturini, L. Golic˘, Z. Trontelj, J. Luz˘nik, Z. Jeglic˘ic˘, and J. Pirnat, Phys. Rev. Lett. 76, 523 共1996兲. 3 D. Arc˘on, P. Cevc, A. Omerzu, and R. Blinc, Phys. Rev. Lett. 80, 1529 共1998兲. 4 Z. Jaglic˘ic˘, Z. Trontelj, J. Luz˘nik, J. Pirnat, and R. Blinc, Solid State Commun. 101, 591 共1997兲. 5 A. Mrzel, P. Cevc, A. Omerzu, and D. Mihailovic´, Phys. Rev. B 53, R2922 共1996兲. 6 T. Kambe, H. Negawa, Y. Nogami, and K. Oshima, Synth. Met.

R865

103, 2424 共1999兲. L. Barbedette, S. Lefrant, T. Yildirim, and J. E. Fischer, in Physics and Chemistry of Fullerenes and Derivatives, edited by H. Kuzmany, J. Fink, M. Mehring, and S. Roth 共World Scientific, London, 1995兲, p. 460. 8 N. Koga and K. Morokuma, Chem. Phys. Lett. 196, 191 共1992兲. 9 T. Matsuo, H. Suga, W. I. F. David, R. M. Ibberson, P. Bernier, A. Zahab, C. Fabre, A. Rassat, and A. Dworkin, Solid State Commun. 83, 711 共1992兲. 10 T. Kawamoto, Solid State Commun. 101, 231 共1996兲. 11 D. Mihailovic´, D. Arc˘on, P. Venturini, R. Blinc, A. Omerzu, and P. Cevc, Science 268, 400 共1995兲. 7